Syed Ahmed is SVP of Engineering at Act-On Software, a leading marketing automation platform.
Artificial intelligence (AI) is always a hot topic in tech. As far back as the 1950s, American movies have portrayed the imaginary thrills and perils of AI. The original 1951 film The Day the Earth Stood Still, for example, depicts a sentient alien robot named Gort that can single-handedly shut down the U.S. military and all human electronic devices.
Although real, modern-day AI innovations like ChatGPT are less authoritarian, people today are understandably still wary of AI’s risks and challenges. However, its true capabilities serve a much less scary purpose.
Let’s start with an AI benefit—cost. The potential for savings through AI-generated coding is a game-changer for the development world. While it may seem like AI is poised to replace human developers, there is much more nuance to this technology than meets the eye.
AI could actually benefit developers everywhere. By automating tedious coding tasks, AI can help elevate everyone’s coding skills to the next level. Instagram cofounder Mike Krieger described GitHub Copilot AI as “the single most mind-blowing application of machine learning” he has ever seen.
According to a 2022 study from Technavio (via Help Net Security), the market for AI is expected to increase by $76.44 billion between 2020 and 2025—with 56% of that growth happening in North America alone. A survey of Stack Overflow users found that machine learning and AI-assisted technologies are the top areas in which developers want more training.
Despite this buzz and potential benefits, some people still feel hesitant about the reliability of AI-generated code. Purdue University professor Saurabh Bagchi told Government Technology it is “not clear how secure or reliable ChatGPT-generated code is.” ChatGPT competitors like Amazon and Meta, meanwhile, continue to double down on developing AI tools (CodeWhisperer and Aroma, respectively).
Throughout these conversations, it’s important to remember AI is not a direct replacement for human developers. Instead, it should be a tool to help streamline coding processes and save time. By automating certain tasks, developers can focus on more important coding tasks that require human expertise.
Consider the issue of technical debt, which is a major concern for many IT decision makers. A Software AG survey found that 21% of the average IT budget was spent on resolving technical debt. Unresolved technical debt can have staggering consequences on app performance and security. The longer you avoid it, the worse it gets. By using AI to speed up internal processes, developers can increase the velocity of new product features, ultimately saving time and money in the long run.
The December 2022 Southwest Airlines debacle is a perfect example of why fixing technical debt is important. After nearly two decades of using the same outdated scheduling software, Southwest was unable to keep up with demand when Winter Storm Elliott forced over a week’s worth of flights to be canceled. The decision cost the company more than $800 million. Having more streamlined and efficient tech tools such as AI coding could help companies avoid some of these costs.
Let’s break down an example of how AI-assisted coding works and can provide these benefits. When migrating a legacy codebase to a more modern system, developers can train OpenAI to do a handful of initial conversions as a proof of concept. Then, they can revise the automated code to fit both their personal style and coding standards. This allows developers to be more productive and shift their focus away from manually tidying up legacy code.
Instead of AI like the fictional robot Gort zapping the power from our tools, real AI can put power back in our hands.
Where can you start? Work with your security team on reviewing the legalese surrounding various AI apps to check on the feasibility of incorporating AI into your development process. It’s important to make sure you procure the correct type of professional licenses for your organization due to these tools’ enhanced security features.
Next, be sure to allow ample time for your development team to prompt the AI effectively. Getting used to prompting and training AI toward your functions can take more time and skill than many people expect, so be prepared to invest the time upfront to save more time down the line.
If you’re starting from zero AI, these steps can feel daunting. However, as Steve Jobs once said: “Innovation distinguishes between a leader and a follower.” While AI-paired coding is still in its infancy, it’s exciting to think about the possibilities for leaders to harness it. From this, we can conclude that AI-assisted coding has the potential to revolutionize the development world by automating tedious coding tasks, increasing productivity and, ultimately, saving time and money.
While there are still some concerns about the reliability of AI-generated code, the benefits are clear: By embracing this technology, developers can focus on what they do best while allowing AI to take care of the rest. The dawn of a new era in coding is upon us, and the future looks brighter than ever.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Read the full article here
